REVIEW

Collagen as a double-edged sword in tumor progression

Min Fang & Jingping Yuan & Chunwei Peng & Yan Li

Received: 12 September 2013 /Accepted: 3 December 2013 /Published online: 15 December 2013# The Author(s) 2013. This article is published with open access at Springerlink.com

Abstract It has been recognized that cancer is not merely adisease of tumor cells, but a disease of imbalance, in whichstromal cells and tumor microenvironment play crucial roles.Extracellular matrix (ECM) as the most abundant componentin tumor microenvironment can regulate tumor cell behaviorsand tissue tension homeostasis. Collagen constitutes the scaf-fold of tumor microenvironment and affects tumor microen-vironment such that it regulates ECM remodeling by collagendegradation and re-deposition, and promotes tumor infiltra-tion, angiogenesis, invasion and migration. While collagenwas traditionally regarded as a passive barrier to resist tumorcells, it is now evident that collagen is also actively involvedin promoting tumor progression. Collagen changes in tumormicroenvironment release biomechanical signals, which aresensed by both tumor cells and stromal cells, trigger a cascadeof biological events. In this work, we discuss how collagencan be a double-edged sword in tumor progression, bothinhibiting and promoting tumor progression at different stagesof cancer development.

Keywords Collagen . ECM remodeling . Tensionhomeostasis . Traction force . Tumor progression

AbbreviationsECM Extracellular matrixFACITs Fibril-associated collagens with

interrupted triple helicesMACITs Membrane-associated collagens with

interrupted triple helices

MULTIPLEXINs Multiple triple-helix domains andinterruptions

BM Basement membraneHCC Hepatocelluar carcinomaLOX Lysyl oxidaseBMDCs Bone marrow derived cellsEMT Epithelial–mesenchymal transitionSPARC Secreted protein acidic and rich in cystineMMPs Matrix metalloproteinasesIL-6 Interleukin-6CSF-1 Colony-stimulating factor-1TAMs Tumor associated macrophagesLAIRs Leukocyte-associated Ig-like receptorsITIMs Immunoreceptor tyrosine-based

inhibition motifsVBM Vascular basement membraneNC1 domain Noncollagenous domain 1

Introduction

Cancer is one of the most serious health threats worldwide,with an estimated 12.7 million new cases and 7.6 millioncancer deaths each year [1]. Invasion and metastasis are themost fundamental properties of tumor biology and the rootcauses of cancer death. To tackle this problem, many effortsfocusing on tumor cells have been made over the past century,and some genetic and epigenetic mechanisms have been elu-cidated [2–6]. Currently, with a general consensus on thesignificance of epigenetics, there has been a re-flowering oftheory that cancer is a disease of imbalance, i.e., not merely adisease of rogue cells but the body's mismanagement of thoserogue cells. It has been well documented that tumor microen-vironment plays an important role in tumor progression via theco-evolution of tumor cells and tumor stroma [7–9]. Exploring

M. Fang : J. Yuan :C. Peng :Y. Li (*)Department of Oncology, Zhongnan Hospital of Wuhan University,Hubei Key Laboratory of Tumor Biological Behaviors and HubeiCancer Clinical Study Center, No. 169 Donghu Road, WuchangDistrict, Wuhan 430071, Chinae-mail: liyansd2@163.com

Tumor Biol. (2014) 35:2871–2882DOI 10.1007/s13277-013-1511-7

the complex mechanisms of tumor progression from perspec-tives of tumor stroma has become a new frontier.

Of note is the extracellular matrix (ECM), a major compo-nent of tumor stroma, as a key regulator of cell and tissuefunction. Traditionally, ECM has been regarded primarily as aphysical scaffold that binds cells and tissues together. How-ever, recent studies have shown that ECM also elicits bio-chemical and biophysical signaling [10, 11] that affects celladhesion and migration, tissue morphogenesis and repair,angiogenesis and cancer, and ECM proteolysis is tightly con-trolled in normal tissues but typically deregulated in cancer[8]. As the most abundant constituent of ECM, collagenaccounts for the major function of ECM, and either increased[12] or decreased [13] deposition of collagen can be associat-ed with increased malignancy.

This review summarized the dynamic interplay betweencollagen and tumor cells, focusing on changes in physico-chemico-biological properties of collagen. A new paradigmhas been formulated that the intrinsic biomechanical forces incollagen can modulate ECM molecular conformation, produc-ing either protective or destructivemolecular and cellular eventsduring tumor progression, depending on the stage of cancerdevelopment. Furthermore, the relationship between collagenand immune response and tumor angiogenesis is also explored.

Basic structure and function of collagen

Collagen is abundant in humans accounting for one-third oftotal proteins. The fibrous, structural protein contains threepolypeptide α-chains, displaying a polyproline-II conforma-tion, a right-handed supercoil and a one-residue stagger be-tween adjacent chains [14]. Each polypeptide chain has arepeating Gly–X–Y triplet, and the three polypeptide α-chains in the triple helix held together by inter-chain hydrogenbonds can be identical, but heterotrimeric triple helices aremore prevalent than homotrimeric triple helices. Gauba andHartgerink [15] observed that assembly of heterotrimeric tri-ple helices was based on the 1:1:1 mixture of (ProLysGly)10/(AspHypGly)10/(ProHypGly)10 (Fig. 1). Collagen undergoesextensive posttranslational modifications by hydroxylationand cross-linking reactions in the endoplasmic reticulum priorto triple helix formation [16]. A number of enzymes andmolecular chaperones assist in their correct folding andtrimerisation, including hydroxylases, collagen glycosyltrans-ferases, peptidyl cis–trans isomerase and protein disulphideisomerase [17, 18]. According to the structure properties ofECM, collagens can be categorized into classical fibrillar andnetwork-forming collagen, FACITs (fibril-associatedcollagens with interrupted triple helices), MACITs(membrane-associated collagens with interrupted triplehelices), and MULTIPLEXINs (multiple triple-helix domainsand interruptions) [19]. At least 28 different types of collagens

have been identified in vertebrates [19, 20] (Table 1). Amongthese, type I collagen is the archetypal collagen in that its triplehelix has no imperfections and it has predominant role intissue [16]. Others can have interruptions in the triple helixand do not necessarily assemble (in their own right) intofibrils. For example, MACIT has numerous interruptions inthe triple helix, does not self-assemble into fibrils, and hasroles in cell adhesion and signaling [20]. And Type IV colla-gen is the prototypical network-forming collagen. It forms aninterlaced network at basement membrane (BM), found at thebasal surface of epithelial and endothelial cells and essentialfor tissue polarity [21], where it has an important molecularfiltration function.

ECM remodeling during cancer invasion

During cancer invasion, tumor stroma undergoes constantarchitectural changes, characterized by collagens degrading,re-depositing, cross-linking and stiffening in terms of ECMremodeling, and immune infiltration and re-differentiation ofmonocytes at the invasive front in terms of cellular changes.

Increased deposition and cross-linking of collagens

The ECM scaffold undergoes considerable structural changesduring tumor progression, including increased deposition offibronectin, proteoglycans and collagens I, III and IV, andenhanced matrix cross-linking [22, 23]. Increased ECM de-position and remodeling creates a reorganized microenviron-ment to promote tumor progression by destabilizing cell po-larity and cell–cell adhesion, and augmenting growth factorsignaling [10, 24]. The progressive ECM remodeling pro-duces typical morphological changes characterized by linear-ization of interstitial collagens at tumor invasion front, withsignificant impacts on tumor cell biology including geneexpression, cell differentiation, proliferation, migration andresponses to treatments [10].

Breast cancer is a typical example of these changes. Clini-cians have long recognized the connection between breastdensity and breast cancer risk [25]. Collagen surroundingnormal epithelial structures in breast tissue is typically curlyand smooth. However, parallel with tumor development, col-lagen progressively thickens, linearizes and stiffens whichpromotes metastasis by fostering cells migration into ECM.Indeed, intravital imaging shows that breast cancer cells andleukocytes migrate rapidly along collagen fibers [26]. Cancercells might exploit these remodeled stiff collagens as invasion"highways", analogous to the preferential migration of gliomacells along the matrix associated with blood vessels and rigidmyelin sheath bundles [27]. Our recent study also observedthe linear invasion "highways" in hepatocelluar carcinoma(HCC) (Fig. 2).

2872 Tumor Biol. (2014) 35:2871–2882

Fig. 1 Biosynthesis of collagen. Three polypeptide α-chains each in-cluding an N- and C-terminal propeptides form triple helical structurescalled procollagen triple helix by lysly hydroxylase, protein disulfideisomerase and hydrogen bonds. Neutral strands are stable, but charged

forms are unstable. Tropocollagen triple helix is formed as N- and C-terminal propeptides are converted into N- and C-terminal peptides by N-and C-proteinases. Under lysyl oxidase (LOX) cross-linking and self-assembly, collagen fibers or networks are formed

Table 1 Collagens in vertebratesat a glance

BM basement membrane, FACITfibril-associated collagens withinterrupted triple helices, MACITmembrane-associated collagenswith interrupted triple helices,MULTIPLEXINs multiple triple-helix domains and interruptions

Type Class Distribution

I Fibril Abundant and widespread in non-cartilaginous connective tissue:dermis, bone, tendon, ligament

II Fibril Cartilage, vitreous

III Fibril Co-distribution with collagen I: skin, blood vessels, intestine

IV Network BM

V Fibril Widespread and co-distribution with collagen I: bone, dermis, cornea, placenta

VI Network Widespread: muscle, bone, cartilage, cornea, dermis

VII FACIT Dermis, bladder

VIII Network Widespread: dermis, brain, heart, kidney

IX FACIT Co-distribution with collagen II: cartilage, cornea, vitreous

X Network Hypertrophic cartilage

XI Fibril Co-distribution with collagen II: cartilage, intervertebral disc

XII FACIT Co-distribution with collagen I: dermis, tendon

XIII MACIT Endothelial cells, dermis, eye, heart

XIV FACIT Widespread and co-distribution with collagen I: bone, dermis, cartilage

XV MULTIPLEXIN Located between collagen fibrils that are close to BM, capillaries, testis,kidney, heart

XVI FACIT Integrated into collagen fibrils and fibrillin-1 microfibrils, dermis, kidney

XVII MACIT Hemidesmosomes in epithelia

XVIII MULTIPLEXIN Associated with BM, liver

XIX FACIT Rare, localized to BM

XX FACIT Widespread: cornea (chick)

XXI FACIT Widespread: stomach, kidney

XXII FACIT Tissue junctions

XXIII MACIT Limited distribution: heart, retina

XXIV Fibril Shares sequence homology with the fibril-forming collagens: bone, cornea

XXV MACIT Brain, heart and testis

XXVI FACIT Testis and ovary

XXVII Fibril Shares sequence homology with the fibril-forming collagens: cartilage

XXVIII Network A component of the BM around Schwann cells, dermis, sciatic nerve

Tumor Biol. (2014) 35:2871–2882 2873

Protease-dependent collagen cross-linking based on LOX

Protease dependent-ECM remodeling is predominantly cata-lyzed by enzymes such as lysyl oxidase (LOX) [28], synthe-sized by either stromal cells during early stages of carcino-genesis, or tumor cells during late stages of tumor progressionin response to hypoxia [29]. As shown in Fig. 3, LOX,secreted by prominent central hypoxic tumor cells, cancrosslink collagens and elastin, thereby increasing insolublematrix deposition and tissue stiffness (Table 2). IncreasedECM stiffness can activate integrins [30], enhance tumor celladhesion and migration [31–33]. LOX is essential in drivingtumor cells escape from primary site, extravasation andgrowth at secondary sites during metastasis [29, 34]. It isreported that LOX can be disseminated into distal targetorgans via circulation to mobilize bone marrow derived cells(BMDCs) to distal sites, and to create pre-metastatic niche[29], as evidenced by consistent correlation between increasedLOX expression and higher cancer metastasis risk [34]. Fur-thermore, increased LOX expression is associated with earlystromal reaction in breast cancer, and reactive fibrosis at theinvasive front of infiltrating tumors also releases high levels ofLOX [33]. The secreted LOX acts on collagen and increasesintegrin activity, stimulating tumor cells to stretch pseudopo-dia protrusions with increased actin polymerization, focaladhesion formation, resulting in the enhancement ofactomyosin- and cytomyosin-dependent cell contractility andmigration, leaving behind remodeled matrix tracks as the"metastasis highway" for tumor cells to travel. In more ag-gressive and poorly differentiated tumors, LOX also inducesepithelial–mesenchymal transition (EMT) and promotes met-astatic dissemination by facilitating tumor cells invasion intovascular system (intravasation) [35].

Protease-independent ECM stiffening

Non-enzymatic collagen cross-linking, such as glycation andtransglutamination or increased biglycan and fibro-modulinproteoglycan deposition, can also stiffen matrix [36]. Suchprotease-independent ECM stiffening could be divided intoseveral models. One kind is the excessive deposition of pro-teoglycans. It could contribute to fibrosis by parenchyma stiff-ening in injured lungs [37], and be accompanied with elevatedrisk of developing cancers in diabetic patients with inappropri-ate glycation-mediated cross-linking [38–40]. Another processis fibronectin-mediated collagen reorganization [48]. The size,density and rigidity of fibronectin in vivo influence function ofcollagen, and dynamic and reciprocal interactions betweencollagen and fibronectin likely induce tumor progression[41]. Indeed, fibronectin has been implicated as early step ofcancer metastasis [42]. And secreted protein acidic and rich incystine (SPARC), a highly conserved, multi-functional glyco-protein, produced both by cancer cells and stromal cells, canalso be involved with such protease-independent ECM stiffen-ing models. It could participate in ECM organization and bindto type I and IV collagen [13] and also suppress or promoteprogression of cancers depending on interactions at cell-matrixand tumor-stroma surface [13, 43–46].

Increased degradation of collagen

Collagen in tissues has been traditionally regarded as merely aphysical barrier against cancer invasion and tumor cells mi-gration [47, 48]. The prerequisite for tumor invasion is colla-gen degradation [49], for which matrix metalloproteinases(MMPs) play an important role [8, 50], with direct causativeeffects on tumor growth, invasion and angiogenesis [51–63],

Fig. 2 Type IV collagenexpression demonstrated byquantum dot-525 (green). aAbundant type IV collagenfragments stochasticallydistributed in tumor tissues. bRich type IV collagen in tumourstroma aligning with tumor nests.c , d Different characteristicsbetween HCC (red star) and livecirrhosis (LC) tissues. Redarrowheads show stiff type IVcollagen at interface of livercirrhosis and tumor nests. Redarrows indicate the linearinvasion "highways" for tumorcells escape. Scale bar=50 μm

2874 Tumor Biol. (2014) 35:2871–2882

by a host of mechanobiological mechanisms such as to de-grade collagen paving a potential tunnel for escaping tumorcells [8], to disturb tumor microenvironment producing addi-tional mechanical force to induce EMT, and to expose activesites on collagen to recruit monocytes leading to a cascade of

innate immuno-inflammatory reactions [54]. However, thereis diversity of tumor invasion mechanisms, in which collagendegradation plays different roles. In single cell/amoeboid mi-gration, cells tend to migrate in the absence of proteolyticECM breakdown by adapting their shape to and squeezing

Fig. 3 The role of LOX in tumor progression both in situ and distalorgans. With tumor growth beyond 2 mm in diameter, prominent centralhypoxia induces tumor cells to secrete LOX into tumormilieu. On the onehand, LOX-mediated type IV collagen cross-linking leads to ECM depo-sition and subsequent tissue stiffness, driving malignant progressionpredominantly by altering integrin focal adhesions and actomyosin- andcytoskeletal-dependent cell contractility. Tumor cells stretch pseudopodia

protrusions with increased actin polymerization, focal adhesion formationand focal adhesion kinase that can in turn enhance tumor cells prolifera-tion, migration, invasion, and perhaps tumor angiogenesis. On the otherhand, LOX is disseminated into target organs (lung in this illustration) viacirculation and deposits at terminal bronchioles and distal alveoli. Thedeposited LOX can crosslink type I and IV collagens to remodel ECM forrecruiting BMDCs, so as to form the pre-metastatic niche

Table 2 Up-regulated expression of LOX in tumor tissues

Cancer type Results Function References

Breast cancer 10-year DMSFa low10-year OSb low

Activate HIF1-Akt pathway; mediate hypoxic control of metastasis; regulateactin filament formation; contribute to mechanotransduction-mediatedregulation of TGF-β signaling; recruit BMDCs to form the pre-metastatic niche

[29, 34, 36]

Colorectal cancer / Correlated with absence of lymphovascular invasion; activate PI3K–Aktpathway to up-regulate HIF-1α protein synthesis

[37, 38]

Head and neck squamouscell carcinoma

5-year OSc low Strongly associated with increased metastasis, progression and death [39]

Lung adenocarcinoma 5-year OS low ECM remodeling; associated with advanced stage and metastasis [40, 41]

Oral and oropharyngealsquamous cell carcinoma

10-year OS low Independent prognostic biomarker and predictor of lymph node metastasis [42]

a 10-year DMSF: 10-year distant metastasis free survivalb 10-year OS low: 10-year overall survivalc 5-year OS low: 5-year overall survival

Tumor Biol. (2014) 35:2871–2882 2875

through tissue gaps and trails. In mesenchymal migration,invading cells adopt spindle-shaped, elongated morphologywith focalized cell-matrix adhesions containing multi-molecular integrin clusters and proteolytic activity towardECM substrates. Focalized proteases on the cell's surfacegenerate small microtracks through which subsequent cellscan follow. In collective invasion, one or several leader cellswith mesenchymal characteristics, such as fibroblasts, formthe tip of multi-cellular strands and generate forward tractionand pericellular proteolysis toward the tissue structure [49].

Recent studies have gained new insights into the functionof MMPs with a new paradigm for mechanobiological mech-anisms in tumor invasion [55], which will be described indetail below.

Reversible changes of BM: opening the doorfor non-proteolytic ECM migration

At initial phase of tumor invasion, BM is breached as tumorcells invade into interstitial tissue and colonize distant organs.Although proteolysis-dependent collagen degradation is im-portant for this process, there is much evidence showing thatproteolysis is dispensable in the BM transmigration events[56, 57].

In this process, the putative mechanism underlying revers-ible BM remodeling could rely on a precedent wherein endo-thelial cells function as the gatekeeper of transmigration byflexibility on BM [58]. That is, endothelial cells can generatetraction force in response to signals during tumor cell–endo-thelial cell adhesive interactions, which regulate collagenstructure and organization. Indeed, BM has been proposed todisplay thixotropic properties: an increasing force generates achange in BM viscosity, altering BM permeability to macro-molecules and perhaps even cells [59]. So, in such a scenario,reversible disruptions of type IV collagen quaternary interac-tions are required for "closing and opening" BM, permittingnon-proteolytic transmigration to occur without enzymaticdegradation [60]. In this manner, a reversible system mightbe envisioned wherein tumor cells or others such as mono-cytes can be accommodated to pass through. Thus, collagencan regulate immune cell infiltration into tumors. Hence,cooperation between traction forces and the activity of cellsurface enzymes on either the endothelial cell or tumor cellsthemselves theoretically would enable the reversible openingand closing of BM [61, 62].

Thus, collagen increase and decrease are both involved intumor progression, and these two processes are coordinatedreciprocally to promote tumor invasion and metastasis. First,LOX mediated collagen cross-linking can recruit some stro-mal cells to adhere and secrete moreMMPs. Then, MMPs candegrade collagen to expose active sites to generating a pro-tumorigenic microenvironment to facilitate tumor progres-sion. Although direct mechanisms have not been elucidated,

MMPs are associated with LOX expression in breast cancer[29]. LOX mediated collagen cross-linking seems to functionin synergy with MMPs, which may lead to ECM remodelingfavoring tumor progression.

Tension homeostasis and tumor progression

Force is essential for normal tissue-specific development, inwhich it regulates cell survival, growth and migration, andorchestrates tissue organization and function. Increased matrixcross-linking and ECM protein deposition or parallel reorien-tation of matrix collagen can stiffen tissue locally to altersurrounding cells growth or drive cell migration. Loss oftissue homeostasis and mechanoreciprocity is a hallmark ofdisease. Although much is known about biochemical path-ways that direct cell behavior, by comparison, little is knownabout how force regulates cell fate and tissue phenotype. Twoprimary cellular mechanisms involved in tumor invasion andmigration are cellular physical rearrangement and reorienta-tion of collagen by traction forces generated by epithelial cells[63, 64], the consequence of EMT and cellular catabolism ofECM by enzymatic cleavage of collagen [65–67]. Both can beregulated by tissue tension. Here we take a multiscale ap-proach to describe tension homeostasis changes present attumor–stroma interface, ranging from the molecular level(collagen and specific enzymes secreted by tumor and stromalcells involved in collagen reorganization) and the cellularlevel (tumor and stromal cells) to the structural level (ECM)[65, 68, 69].

EMT process

Interactions between tumor cells and their surrounding ECMare recognized as primary forces driving the EMT process[68]. Imbalanced biomechanical force at tumor–stroma inter-face is the key trigger initiating EMT, and ultimately leads totumor cells escaping [55]. Tumor cells dynamically adapt tothe force (Fig. 4) by changing their behaviors and remodelingtheir surrounding microenvironment. As the tumor mass ex-pands at early stages of the invasion, collagen in stroma willrealign and stretch perpendicular to the mass to resist tumorexpansion and enzymatic degradation. Thus, tumor cells mustovercome increased collagen alignment and density beforeinvasion and migration. With the tumor mass expanding,stress on ECM increases correspondingly, until reaching acritical point, termed as the biomechanical trigger [62], whichcan be sensed by both tumor and stromal cells through mech-anoreceptors. In turn these cells exert actomyocin- andcytoskeletal-dependent traction forces on ECM [70–72].Eventually, tumor and stromal cells deform as consequencesto the altered tissue tension [73, 74], the expanding tumormass [24], matrix stiffening [10], and increased interstitial

2876 Tumor Biol. (2014) 35:2871–2882

pressure due to a leaky vasculature and poor lymphatic drain-age, initiating EMT [75]. These deformed cells acquire a morespindle-like fibroblastic morphology, less adhesive properties,enhanced motility and invasive behavior.

Tension as a regulator for MMPs function

ECM degradation is necessary for tumor invasion. However,collagen as one of the most abundant components of ECM isthe largest, strongest and most difficult to penetrate by tumorcells due to limited degradation by only a few MMPs. Andalso collagen undergoes remodeling and reorientation in re-sponse to tumor mass expansion, with a concomitant increasein collagen density and mechanical tension to constrain tumorexpansion [70, 76]. Re-constructed collagen can prevent themfrom being degraded by MMPs. As shown in Fig. 5, in normaltissues,MMPs can access into collagen to degrade (Fig. 5a1–a2).However, in tumor tissues, function ofMMPs is regulated by thetumor microenvironment, here accenting the coordination ofstromal cells, tumor cells and collagen as well: (1) initially,increased tension in ECM is protective, it makes the collagenstretched, in that it inhibits MMP-related collagen cleavage ascollagen undergoes molecular conformation changes, such thatthe enzymes no longer have access to the cleavage site oncollagen as the binding sites hidden (Fig. 5b1–b2); (2) with thetumor extensively expansion, tumor cells and stromal cells sensehigh tension in ECM so as to exert traction force on collagens tomake them deform through integrin binding (Fig. 5c1–c2); (3) astension exerting on cells increases with further tumor massexpansion until a critical point, rather than being protective for

ECM degradation, increasing tension becomes a key biome-chanical trigger for tumor and stroma cells to remodel theECM. Via surface integrin receptors, cancer cells can sensemechanical signals of increasing tension in the microenviron-ment, and respond by increasing their traction forces on ECM,resulting in collagen triple helix separation (Fig. 5d1–d2); (4)thus, entrance hole for MMPs is opened by destabilizingcollagen triple helix and unwinding collagen molecule(single α-chain) [62], accompanied with tension decrease inECM (Fig. 5e1–e2).

Thus, during tumor progression, ECM remodeling contrib-uting to tension changes is important for both triggering EMTand regulating MMPs. As the tumor expands, the tension inECM will increase correspondingly, until reaching a criticalpoint [55]. This may be a turning point for most tumors andbefore reaching this turning point, tumor progresses slowly,but beyond this turning point, tumor progression isaccelerated.

Collagen as a regulator for tumor associated immuneinfiltration

Collagen is not just a passive player during tumor progression.Recent experimental developments point to a far more com-plex role for these structure proteins. A variety of immunecells are present in cancers and many of these accumulate andmigrate within regions of dense collagen [26, 77, 78].

For example, macrophages in and around tumor nests can,in principle, either promote or inhibit tumor progression [79],

Fig. 4 Force applied to deformand influence the biologicalbehavior of tumor cells. Tissuemicroenvironment can exertthree forms of force on tumorcells, including shear stress,compressive stress andtensile stress

Tumor Biol. (2014) 35:2871–2882 2877

and collagen plays a crucial role in regulating the balancebetween the tumor-inhibiting and promoting effects of mac-rophages. It is reported that culturing macrophages on type Icollagen reduces their cytotoxicity against tumor cells [80],suggesting that collagen inhibits the differentiation of themacrophages to the tumoricidal M1-like type. The possibilitythat collagen scaffolds can regulate macrophages polarizationis further supported by the increase in pro-tumorigenic, M2-like macrophages observed in tumors of Sparc−/− mice withabnormal collagen scaffolds [13]. The principal factors behindthis transformation appear to be interleukin-6 (IL-6) andcolony-stimulating factor 1 (CSF-1) [81]. And it is also knownthat collagen degradation products serve as chemotactic stim-uli for monocytes [82, 83]. As collagen is degraded duringmetastasis, the resulting collagen fragments may recruit tumorassociated macrophages (TAMs). These TAMs are abundantin most solid tumors [84], predicting poor prognosis [85]. Asshown in Fig. 6, monocytes are recruited into collagens deg-radation areas accompanied with MMPs release by tumorcells [85]. After releasing soluble CSF-1, monocytes differen-tiate into TAMs promoting tumor growth and metastasis [85].

Similarly, IL-6 will inhibit monocytes differentiation intodendritic cells, directing immune responses against tumorcells [85]. TAMs themselves express factors in response totumor progression, promoting tumor angiogenesis, invasionand intra-and extravasation [86, 87]. For example, IL-1β, aspecial form of IL-1, can stimulate expression of VEGF andTNFα to promote tumor angiogenesis and adhesion mole-cules, including intercellular-adhesion molecule 1 (ICAM-1),vascular cell adhesion molecule 1 (VCAM-1) and E-selectin,to enhance invasion. And IL-1β can also activate MMPs todegrade collagen [88, 89]. Thus, all these have uncovered asignificant positive feedback in immune responses to cancer-related collagen degradation and indicate the link betweendegraded type IV collagen and tumor progression [90].

In addition, ECM stiffness could influence T cell activationvia integrin-mediated adhesions assembly promotion [10, 90]and collagen-mediated activation of leukocyte-associated Ig-like receptors (LAIRs). LAIRs are highly expressed on mostimmune cells and can through their immunoreceptor tyrosine-based inhibition motifs (ITIMs) inhibit immune cell activation[91]. Although it is not clear whether LAIRs and integrins

Fig. 5 A paradigm for how tumor and stromal cells interact to degradeECM and change tensions for tumor invasion. a1 Dormant tumor cellswithout tension force; a2 collagen relaxed and elastic. b1 With tumorgrowth, low tension force exerts on collagen which stretches accordingly;b2 Entrance hole for MMPs-dependent cleavage is closed as collagenstretches. c1 As the tumor continues to expand, increasing tension forcetransmits signals to both tumor and stromal cells to remodel ECM in order

to reduce tension force. Tumor and stromal cells undergo EMT processwhich in turn increase their traction force; c2 collagen bends and changesconformational structures correspondingly. d1 High traction forceexerted by cells destabilize the stroma; d2 tumor and stromal cells attachto collagen and unwind triple helix, exposing sites for cleavage byMMPs. e1 Tumor invasion and metastasis occur with degradation ofcollagen; e2 MMPs enter into triple helix to cleave α-chains

2878 Tumor Biol. (2014) 35:2871–2882

cooperate, activation of LAIRs is a plausible mechanismwhereby high levels of deposited collagen lead to inhibitionof an anti-tumor immune response.

As collagen influences immune cell infiltration, immunecells also influence collagen architecture. Macrophages regu-late mammary epithelial invasion during tissue development[92]. This may in part be achieved through their ability toinitiate the remodeling and reorganization of collagen sur-rounding the developing epithelium [93], and secretion of arepertoire of soluble factors such as MMPs. Macrophages canalso take up collagen for intracellular degradation via bindingto the glycoprotein [94, 95].

Collagen and tumor angiogenesis

Angiogenesis, a specialized form of branching morphogenesiswherein endothelial cells detach themselves from the existingvasculature, invade surrounding tissues, and reorganize intopatent tubules [86, 96], is vital for tumor growth and metasta-sis. Tumor angiogenesis is characterized by the secretion ofmultiple pro-angiogenic factors to trigger the angiogenicswitch resulting in the development of a structurally andfunctionally abnormal vasculature.

Collagens are essential for tumor angiogenesis. Inhibitionof collagen metabolism has been demonstrated to have anti-angiogenic effects [97], confirming that blood vessel forma-tion and survival are indispensably connected with propercollagen synthesis and deposition at BM [97]. Interactionsbetween endothelial cells and ECM, in particular collagen

IV in the vascular basement membrane (VBM) play key rolesin regulating angiogenesis [21]. For instance, type IV collagencould modulate (promote/inhibit) endothelial cells growth andproliferation [98]. The in vitro endothelial cells culture exper-iments have shown that triple-helical fragments of type IVcollagen could stimulate endothelial-cell adhesion and migra-tion as active as intact type IV collagen, while thenoncollagenous domain 1 (NC1 domain) of type IV collagenalone is insufficient to mediate endothelial-cell migration [99,100]. And a further research of NC1 domains indicated thatthey are typical anti-angiogenic molecules, which could in-hibit endothelial cells migration, proliferation and tube forma-tion by competing with intact type IV collagen for bindingintegrin [101]. In addition, studies on in situ carcinoma dem-onstrated that MMP-mediated degradation of BM could ex-pose cryptic domains of type IV collagen with pro-angiogenicactivity, in the early stage of local tumor progression [102,103], and generate type IV collagen fragments with anti-angiogenic activity, such as arrestin, canstatin and tumstatinin the late stage [104–106]. Thus the structural integrity ofcollagen IV is of utmost importance for tumor angiogenesis[21].

Conclusions and future perspectives

Over the last 10 years, cancer research has been increasinglyshifted to the tumor microenvironment. In particular, ECM,the intermediary between biomechanics and tumor biology,can mediate dual roles as tumor suppressors at the early stages

Fig. 6 Collagen regulates tumor associated immune infiltration. MMP-dependent collagen fragments can recruit monocytes and further promotethem to differentiate into TAMs with the help of CSF-1. TAMs

themselves secret factors responsible for tumor progression, includingtumor angiogenesis. Meanwhile, they themselves can activate MMPs todegrade collagens

Tumor Biol. (2014) 35:2871–2882 2879

but paradoxically as tumor promoters at the later stages oftumor progression. Current researches of ECM focus on bio-chemical mechanisms associated with tumor progression,namely the intracellular pathways of signal transduction fromthe ECM to the nucleus (outside-in signaling), and the cellularmetabolic responses for synthesizing proteinases to degrade(inside-out signaling) [62]. However, little attention has beenpaid to the dynamic changes of ECM biomechanics accom-panying these biochemical events. With the increasing appre-ciation that biomechanical forces are crucial determinants fortissue development, cell differentiation and homeostasis, it isreasonable to conclude that loss of the ability to sense, respondand adapt appropriately to such biomechanical forces, on thepart of tumor cells and stromal cells, contributes to tumorprogression. Therefore, collagen as the most important archi-tecture of ECM to generate these biomechanical forces, is nolonger considered as a static and passive background uponwhich metastasis takes place. To elucidate how the changes incollagen structure and the related biomechanical forces tomodulate tumor invasion and metastasis, thus decipheringthe "collagen code" in cancer progression, is an intriguingfield for intensive investigation.

Acknowledgments This work was supported by the National NaturalScience Foundation of China (81171396, 81230031/H18), the NationalScience and Technology Major Project (2012ZX10002012-12) and theNational University Students Innovation Training Project of China(111048673).

Conflicts of interest The authors declare no conflict of interest.

Open Access This article is distributed under the terms of the CreativeCommons Attribution License which permits any use, distribution, andreproduction in any medium, provided the original author(s) and thesource are credited.

References

1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Globalcancer statistics. CA Cancer J Clin. 2011;61:69–90.

2. Egeblad M, Nakasone ES, Werb Z. Tumors as organs: complextissues that interface with the entire organism. Dev Cell. 2010;18:884–901.

3. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D,Gronroos E, et al. Intratumor heterogeneity and branched evolutionrevealed by multiregion sequencing. N Engl J Med. 2012;366:883–92.

4. Rosner M, Hengstschlager M. Targeting epigenetic readers in can-cer. N Engl J Med. 2012;367:1764–5.

5. Dawson MA, Kouzarides T, Huntly BJ. Targeting epigeneticreaders in cancer. N Engl J Med. 2012;367:647–57.

6. Giovannetti E, Codacci-Pisanelli G, Peters GJ. TFAP2E-DKK4 andchemoresistance in colorectal cancer. N Engl J Med. 2012;366:966.

7. GarciaMG, Bayo J, BolontradeMF, Sganga L,Malvicini M, AlanizL, et al. Hepatocellular carcinoma cells and their fibrotic microen-vironment modulate bone marrow-derived mesenchymal stromalcell migration in vitro and in vivo. Mol Pharm. 2011;8:1538–48.

8. Page-McCaw A, Ewald AJ, Werb Z. Matrix metalloproteinases andthe regulation of tissue remodelling. Nat RevMol Cell Biol. 2007;8:221–33.

9. Yu D, Zhuang L, Sun X, Chen J, Yao Y, Meng K, et al. Particulardistribution and expression pattern of endoglin (CD105) in the liverof patients with hepatocellular carcinoma. BMC Cancer. 2007;7:122.

10. Paszek MJ, Zahir N, Johnson KR, Lakins JN, Rozenberg GI, GefenA, et al. Tensional homeostasis and the malignant phenotype.Cancer Cell. 2005;8:241–54.

11. Butcher DT, Alliston T, Weaver VM. A tense situation: forcingtumour progression. Nat Rev Cancer. 2009;9:108–22.

12. Levental KR, Yu H, Kass L, Lakins JN, Egeblad M, Erler JT, et al.Matrix crosslinking forces tumor progression by enhancing integrinsignaling. Cell. 2009;139:891–906.

13. Arnold SA, Rivera LB, Miller AF, Carbon JG, Dineen SP, Xie Y,et al. Lack of host SPARC enhances vascular function and tumorspread in an orthotopic murine model of pancreatic carcinoma. DisModel Mech. 2010;3:57–72.

14. Brodsky B, Persikov AV. Molecular structure of the collagen triplehelix. Adv Protein Chem. 2005;70:301–39.

15. Gauba V, Hartgerink JD. Self-assembled heterotrimeric collagentriple helices directed through electrostatic interactions. J AmChem Soc. 2007;129:2683–90.

16. Boot-Handford RP, Tuckwell DS. Fibrillar collagen: the key tovertebrate evolution? A tale of molecular incest. BioEssays.2003;25:142–51.

17. Vanacore R, Ham AJL, Voehler M, Sanders CR, Conrads TP,Veenstra TD, et al. A sulfilimine bond identified in collagen IV.Science. 2009;325:1230–4.

18. Myllyharju J, Kivirikko KI. Collagens, modifying enzymes andtheir mutations in humans, flies and worms. Trends Genet.2004;20:33–43.

19. Shoulders MD, Raines RT. Collagen structure and stability. AnnuRev Biochem. 2009;78:929–58.

20. Kadler KE, Baldock C, Bella J, Boot-Handford RP. Collagens at aglance. J Cell Sci. 2007;120:1955–8.

21. Kalluri R. Basement membranes: structure, assembly and role intumour angiogenesis. Nat Rev Cancer. 2003;3:422–33.

22. Zhu GG, Risteli L, Makinen M, Risteli J, Kauppila A, Stenback F.Immunohistochemical study of type I collagen and type I pN-collagen in benign and malignant ovarian neoplasms. Cancer.1995;75:1010–7.

23. LesniakM, Huijbers IJ, IravaniM, Popov S, Robertson D, Al-SarrajS, et al. A role for fibrillar collagen deposition and the collageninternalization receptor Endo180 in glioma invasion. PLoS ONE.2010;5:e9808.

24. Paszek MJ, Weaver VM. The tension mounts: mechanics meetsmorphogenesis and malignancy. J Mammary Gland Biol Neoplasia.2004;9:325–42.

25. Wolfe JN. Risk for breast cancer development determined by mam-mographic parenchymal pattern. Cancer. 1976;37:2486–92.

26. Wyckoff JB, Wang Y, Lin EY, Li JF, Goswami S, Stanley ER, et al.Direct visualization of macrophage-assisted tumor cell intravasationin mammary tumors. Cancer Res. 2007;67:2649–56.

27. Giese A, Kluwe L, Laube B, Meissner H, Berens ME, Westphal M.Migration of human glioma cells on myelin. Neurosurgery.1996;38:755–64.

28. Xiao Q, Ge G. Lysyl Oxidase, Extracellular matrix remodeling andcancer metastasis. Cancer Microenviron. 2012;5:261–73.

29. Erler JT, Bennewith KL, Cox TR, Lang G, Bird D, Koong A, et al.Hypoxia-induced lysyl oxidase is a critical mediator of bone mar-row cell recruitment to form the premetastatic niche. Cancer Cell.2009;15:35–44.

30. Tzima E, del Pozo MA, Shattil SJ, Chien S, Schwartz MA.Activation of integrins in endothelial cells by fluid shear stress

2880 Tumor Biol. (2014) 35:2871–2882

mediates Rho-dependent cytoskeletal alignment. EMBO J.2001;20:4639–47.

31. Galbraith CG, Yamada KM, Sheetz MP. The relationship betweenforce and focal complex development. J Cell Biol. 2002;159:695–705.

32. Riveline D, Zamir E, Balaban NQ, Schwarz US, Ishizaki T,Narumiya S, et al. Focal contacts as mechanosensors: externallyapplied local mechanical force induces growth of focal contacts byan mDia1-dependent and ROCK-independent mechanism. J CellBiol. 2001;153:1175–86.

33. Erler JT, Giaccia AJ. Lysyl oxidase mediates hypoxic control ofmetastasis. Cancer Res. 2006;66:10238–41.

34. Erler JT, Bennewith KL, NicolauM, Dornhöfer N, Kong C, Le Q-T,et al. Lysyl oxidase is essential for hypoxia-induced metastasis.Nature. 2006;440:1222–6.

35. Polyak K, Weinberg RA. Transitions between epithelial and mes-enchymal states: acquisition of malignant and stem cell traits. NatRev Cancer. 2009;9:265–73.

36. Avery NC, Bailey AJ. The effects of the Maillard reaction on thephysical properties and cell interactions of collagen. Pathol Biol(Paris). 2006;54:387–95.

37. Ebihara T, Venkatesan N, Tanaka R, Ludwig MS. Changes inextracellular matrix and tissue viscoelasticity in bleomycin-induced lung fibrosis. Temporal aspects. Am J Respir Crit CareMed. 2000;162:1569–76.

38. Robins SP, Milne G, Duncan A, Davies C, Butt R, Greiling D, et al.Increased skin collagen extractability and proportions of collagentype III are not normalized after 6 months healing of human exci-sional wounds. J Investig Dermatol. 2003;121:267–72.

39. Susic D. Cross-link breakers as a new therapeutic approach tocardiovascular disease. Biochem Soc Trans. 2007;35:853–6.

40. Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ. Diabetesmellitus as a predictor of cancer mortality in a large cohort of USadults. Am J Epidemiol. 2004;159:1160–7.

41. Velling T, Risteli J, Wennerberg K, Mosher DF, Johansson S.Polymerization of type I and III collagens is dependent on fibro-nectin and enhanced by integrins alpha 11beta 1 and alpha 2beta 1. JBiol Chem. 2002;277:37377–81.

42. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, CostaC, et al. VEGFR1-positive haematopoietic bone marrow progenitorsinitiate the pre-metastatic niche. Nature. 2005;438:820–7.

43. Gorantla B, Bhoopathi P, Chetty C, Gogineni VR, Sailaja GS,Gondi CS, et al. Notch signaling regulates tumor-induced angio-genesis in SPARC-overexpressed neuroblastoma. Angiogenesis.2012;16:85–100.

44. Chen J,WangM, Xi B, Xue J, He D, Zhang J, et al. SPARC is a keyregulator of proliferation, apoptosis and invasion in human ovariancancer. PLoS ONE. 2012;7:e42413.

45. Zhang Y, Yang B, Du Z, Bai T, Gao YT, Wang YJ, et al. Aberrantmethylation of SPARC in human hepatocellular carcinoma and itsclinical implication. World J Gastroenterol. 2012;18:2043–52.

46. Chlenski A, Cohn SL. Modulation of matrix remodeling by SPARCin neoplastic progression. Semin Cell Dev Biol. 2010;21:55–65.

47. Baba Y, Iyama K-I, Ikeda K, Ishikawa S, Hayashi N, Miyanari N,et al. The expression of type IV collagen α6 chain is related to theprognosis in patients with esophageal squamous cell carcinoma.Ann Surg Oncol. 2007;15:555–65.

48. Sabeh F, Shimizu-Hirota R, Weiss SJ. Protease-dependent versus -independent cancer cell invasion programs: three-dimensionalamoeboid movement revisited. J Cell Biol. 2009;185:11–9.

49. Friedl P, Wolf K. Plasticity of cell migration: a multiscale tuningmodel. J Cell Biol. 2010;188:11–9.

50. Rowe RG, Weiss SJ. Breaching the basement membrane: who,when and how? Trends Cell Biol. 2008;18:560–74.

51. Taniwaki K, Fukamachi H, Komori K, Ohtake Y, Nonaka T,Sakamoto T, et al. Stroma-derived matrix metalloproteinase

(MMP)-2 promotes membrane type 1-MMP-dependent tumorgrowth in mice. Cancer Res. 2007;67:4311–9.

52. Hu M, Yao J, Carroll DK, Weremowicz S, Chen H, Carrasco D,et al. Regulation of in situ to invasive breast carcinoma transition.Cancer Cell. 2008;13:394–406.

53. Devy L, Huang L, Naa L, Yanamandra N, Pieters H, Frans N, et al.Selective inhibition of matrix metalloproteinase-14 blocks tumorgrowth, invasion, and angiogenesis. Cancer Res. 2009;69:1517–26.

54. Parks WC, Wilson CL, Lopez-Boado YS. Matrix metalloprotein-ases as modulators of inflammation and innate immunity. Nat RevImmunol. 2004;4:617–29.

55. Torzilli PA, Bourne JW, Cigler T, Vincent CT. A new paradigm formechanobiological mechanisms in tumor metastasis. Semin CancerBiol. 2012;22:385–95.

56. Friedl P, Wolf K. Tumour-cell invasion and migration: diversity andescape mechanisms. Nat Rev Cancer. 2003;3:362–74.

57. Gadea G, de Toledo M, Anguille C, Roux P. Loss of p53 promotesRhoA-ROCK-dependent cell migration and invasion in 3D matri-ces. J Cell Biol. 2007;178:23–30.

58. Rabodzey A, Alcaide P, Luscinskas FW, Ladoux B. Mechanicalforces induced by the transendothelial migration of human neutro-phils. Biophys J. 2008;95:1428–38.

59. Simpson LO. Basement membranes and biological thixotropy: anew hypothesis. Pathology. 1980;12:377–89.

60. Shintani Y, Fukumoto Y, Chaika N, Svoboda R, Wheelock MJ,Johnson KR. Collagen I-mediated up-regulation of N-cadherin re-quires cooperative signals from integrins and discoidin domainreceptor 1. J Cell Biol. 2008;180:1277–89.

61. Borza DB, Bondar O, Colon S, Todd P, Sado Y, Neilson EG, et al.Goodpasture autoantibodies unmask cryptic epitopes by selectivelydissociating autoantigen complexes lacking structural reinforce-ment: novel mechanisms for immune privilege and autoimmunepathogenesis. J Biol Chem. 2005;280:27147–54.

62. Essex DW, Li M, Miller A, Feinman RD. Protein disulfide isomer-ase and sulfhydryl-dependent pathways in platelet activation.Biochemistry. 2001;40:6070–5.

63. Hoffman BD, Grashoff C, Schwartz MA. Dynamic molecular pro-cesses mediate cellular mechanotransduction. Nature. 2011;475:316–23.

64. Mierke CT, Frey B, Fellner M, Herrmann M, Fabry B. Integrinalpha5beta1 facilitates cancer cell invasion through enhanced con-tractile forces. J Cell Sci. 2011;124:369–83.

65. Wolf K, Wu YI, Liu Y, Geiger J, Tam E, Overall C, et al. Multi-steppericellular proteolysis controls the transition from individual tocollective cancer cell invasion. Nat Cell Biol. 2007;9:893–904.

66. Gonzalez LO, Gonzalez-Reyes S, Junquera S,Marin L, Gonzalez L,Del Casar JM, et al. Expression of metalloproteases and theirinhibitors by tumor and stromal cells in ductal carcinoma in situ ofthe breast and their relationship with microinvasive events. J CancerRes Clin Oncol. 2010;136:1313–21.

67. del Casar JM, Carreno G, Gonzalez LO, Junquera S, Gonzalez-Reyes S, Gonzalez JM, et al. Expression of metalloproteases andtheir inhibitors in primary tumors and in local recurrences aftermastectomy for breast cancer. J Cancer Res Clin Oncol. 2010;136:1049–58.

68. Motrescu ER, Blaise S, Etique N, Messaddeq N, Chenard MP, StollI, et al. Matrix metalloproteinase-11/stromelysin-3 exhibitscollagenolytic function against collagen VI under normal and ma-lignant conditions. Oncogene. 2008;27:6347–55.

69. Vincent T, Neve EP, Johnson JR, Kukalev A, Rojo F, Albanell J,et al. A SNAIL1–SMAD3/4 transcriptional repressor complex pro-motes TGF-beta mediated epithelial–mesenchymal transition. NatCell Biol. 2009;11:943–50.

70. Provenzano PP, Eliceiri KW, Campbell JM, Inman DR, White JG,Keely PJ. Collagen reorganization at the tumor–stromal interfacefacilitates local invasion. BMC Med. 2006;4:38.

Tumor Biol. (2014) 35:2871–2882 2881

71. Provenzano P, Alejandroosorio A, ValhmuW, Jensen K, VanderbyjrR. Intrinsic fibroblast-mediated remodeling of damaged collage-nous matrices in vivo. Matrix Biol. 2005;23:543–55.

72. Shieh AC. Biomechanical forces shape the tumor microenviron-ment. Ann Biomed Eng. 2011;39:1379–89.

73. Bao G, Suresh S. Cell and molecular mechanics of biologicalmaterials. Nat Mater. 2003;2:715–25.

74. Suresh S. Biomechanics and biophysics of cancer cells. ActaBiomater. 2007;3:413–38.

75. Netti PA, Berk DA, Swartz MA, Grodzinsky AJ, Jain RK. Role ofextracellular matrix assembly in interstitial transport in solid tumors.Cancer Res. 2000;60:2497–503.

76. TangX,Kuhlenschmidt TB, Zhou J, Bell P,Wang F, KuhlenschmidtMS, et al. Mechanical force affects expression of an in vitrometastasis-like phenotype in HCT-8 cells. Biophys J. 2010;99:2460–9.

77. Condeelis J, Segall JE. Intravital imaging of cell movement intumours. Nat Rev Cancer. 2003;3:921–30.

78. Egeblad M, Ewald AJ, Askautrud HA, Truitt ML, Welm BE,Bainbridge E, et al. Visualizing stromal cell dynamics in differenttumor microenvironments by spinning disk confocal microscopy.Dis Model Mech. 2008;1:155–67.

79. Pollard JW. Tumour-educated macrophages promote tumour pro-gression and metastasis. Nat Rev Cancer. 2004;4:71–8.

80. Kaplan G. In vitro differentiation of human monocytes. Monocytescultured on glass are cytotoxic to tumor cells but monocytes cul-tured on collagen are not. J Exp Med. 1983;157:2061–72.

81. Menetrier-Caux C, Montmain G, Dieu MC, Bain C, Favrot MC,Caux C, et al. Inhibition of the differentiation of dendriticcells from CD34(+) progenitors by tumor cells: role ofinterleukin-6 and macrophage colony-stimulating factor.Blood. 1998;92:4778–91.

82. Weathington NM, van Houwelingen AH, Noerager BD, JacksonPL, Kraneveld AD, Galin FS, et al. A novel peptide CXCR ligandderived from extracellular matrix degradation during airway inflam-mation. Nat Med. 2006;12:317–23.

83. Postlethwaite AE, Kang AH. Collagen-and collagen peptide-induced chemotaxis of human blood monocytes. J Exp Med.1976;143:1299–307.

84. Normann SJ. Macrophage infiltration and tumor progression.Cancer Metastasis Rev. 1985;4:277–91.

85. Bingle L, Brown NJ, Lewis CE. The role of tumour-associatedmacrophages in tumour progression: implications for new antican-cer therapies. J Pathol. 2002;196:254–65.

86. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases.Nature. 2000;407:249–57.

87. Boudreau N, Myers C. Breast cancer-induced angiogenesis: multi-ple mechanisms and the role of the microenvironment. BreastCancer Res. 2003;5:140–6.

88. Voronov E, Shouval DS, Krelin Y, Cagnano E, Benharroch D,Iwakura Y, et al. IL-1 is required for tumor invasiveness andangiogenesis. Proc Natl Acad Sci U S A. 2003;100:2645–50.

89. Chirivi RG, Chiodoni C, Musiani P, Garofalo A, Bernasconi S,Colombo MP, et al. IL-1alpha gene-transfected human melanomacells increase tumor-cell adhesion to endothelial cells and theirretention in the lung of nude mice. Int J Cancer. 1996;67:856–63.

90. Roth JM, Caunt M, Cretu A, Akalu A, Policarpio D, Li X, et al.Inhibition of experimental metastasis by targeting the HUIV26cryptic epitope in collagen. Am J Pathol. 2006;168:1576–86.

91. Meyaard L. The inhibitory collagen receptor LAIR-1 (CD305). JLeukoc Biol. 2008;83:799–803.

92. Gouon-Evans V, Rothenberg ME, Pollard JW. Postnatal mammarygland development requires macrophages and eosinophils.Development. 2000;127:2269–82.

93. Ingman WV, Wyckoff J, Gouon-Evans V, Condeelis J, Pollard JW.Macrophages promote collagen fibrillogenesis around terminal endbuds of the developingmammary gland. Dev Dyn. 2006;235:3222–9.

94. Atabai K, Jame S, Azhar N, Kuo A, Lam M, McKleroy W, et al.Mfge8 diminishes the severity of tissue fibrosis in mice by bindingand targeting collagen for uptake by macrophages. J Clin Invest.2009;119:3713–22.

95. Leitinger B, Hohenester E. Mammalian collagen receptors. MatrixBiol. 2007;26:146–55.

96. Egeblad M, Werb Z. New functions for the matrix metalloprotein-ases in cancer progression. Nat Rev Cancer. 2002;2:161–74.

97. Maragoudakis ME, Missirlis E, Karakiulakis GD, Sarmonica M,Bastakis M, Tsopanoglou N. Basement membrane biosynthesis as atarget for developing inhibitors of angiogenesis with anti-tumorproperties. Kidney Int. 1993;43:147–50.

98. Madri JA. Extracellular matrix modulation of vascular cell behav-iour. Transpl Immunol. 1997;5:179–83.

99. Herbst TJ, McCarthy JB, Tsilibary EC, Furcht LT. Differentialeffects of laminin, intact type IV collagen, and specific domains oftype IV collagen on endothelial cell adhesion and migration. J CellBiol. 1988;106:1365–73.

100. Tsilibary EC, Reger LA, Vogel AM, Koliakos GG, Anderson SS,Charonis AS, et al. Identification of multifunctional, cell-bindingpeptide sequence from the a1(NC1) of type IV collagen. J Cell Biol.1990;111:1583–91.

101. Sudhakar A, Nyberg P, Keshamouni VG, Mannam AP, Li J,Sugimoto H, et al. Human alpha1 type IV collagen NC1 domainexhibits distinct antiangiogenic activity mediated by alpha1beta1integrin. J Clin Invest. 2005;115:2801–10.

102. Coussens LM, Werb Z. Inflammatory cells and cancer: think differ-ent. J Exp Med. 2001;193:23–6.

103. Xu JS, Rodriguez D, Petitclerc E, Kim JJ, Hangai M, Yuen SM,et al. Proteolytic exposure of a cryptic site within collagen type IV isrequired for angiogenesis and tumor growth in vivo. J Cell Biol.2001;154:1069–79.

104. Kim YM, Jang JK, Lee OH, Yeon J, Choi EY, Ki KW, et al.Endostatin inhibits endotheial and tumor cellular invasion byblocking the activation and catalytic activity of matrix metallopro-teinase 2. Cancer Res. 2000;60:5410–3.

105. Kamphaus GD, Colorado PC, Panka DJ, Hopfer H, Ramchandran R,Torre A, et al. Canstatin, a novel matrix-derived inhibitor ofangiogenesis and tumor growth. J Biol Chem. 2000;275:1209–15.

106. Petitclerc E, Boutaud A, Prestayko A, Xu JS, Sado Y, Ninomiya Y,et al. New functions for non-collagenous domains of human colla-gen type IV: novel integrin ligands inhibiting angiogenesis andtumor growth in vivo. J Biol Chem. 2000;275:8051–61.

2882 Tumor Biol. (2014) 35:2871–2882